The following MARS requests can be used as a tutorial. The links into the MARS Catalogue might be helpful to create a valid retrieve script. However, the dates might need to be adapted.

As one can very easily retrieve large amounts of data it is best to start mars in a location where plenty of disk space is available, e.g. in $SCRATCH:

cd $SCRATCH

The solutions below provide valid MARS requests which can be passed to the mars command as

% mars < myrequestfile
                             or
% mars myrequestfile

Simple retrieves

Retrieve 1000,850,700 and 500 hPa temperatures from the 15th of February 1999 12 UTC Analysis  > see in MARS Catalogue 

retrieve,
    class    = od,
    type     = an, 
    expver   = 1,
    date     = 19990215,
    time     = 12,
    param    = 130.128, 
    levtype  = pressure level,
    levelist = 1000/850/700/500,
    target   = "t.grb"

Retrieve wind components valid for 5 days ago at 12:00 and 18:00 from forecast made 7 days ago at 12 UTC at all 137 model levels  > see in MARS Catalogue

retrieve,
    type     = forecast, 
    param    = 131/132, 
    levtype  = model level,
    levelist = 1/to/137,
    date     = -7,
    time     = 12,
    step     = 48/54, 
    target   = "wind"

Retrieve geopotential and surface pressure from analysis produced 2 days ago, analysis cycles (00,06,12,18)  > see in MARS Catalogue

retrieve,
    class   = od,
    type    = analysis,
    levtype = surface,
    date    = -2,
    time    = 00/06/12/18,
    param   = 129.128/134.128,
    target  = "all.an"

Retrieve surface pressure and surface long-wave (thermal) radiation downwards at every 12-hour time-step for the 12Z forecast from 12 days ago  > see in MARS Catalogue

retrieve,
    type    = forecast,
    param   = 134.128/175.128,
    date    = -12,
    step    = 12/to/240/by/12,
    time    = 12,
    levtype = sfc,
    target  = "sfc.grb"

Retrieve the operational monthly mean model level temperature (all levels) from the 06:00 analysis for February 1997  > see in MARS Catalogue

retrieve,
    class    = od,
    stream   = mnth,
    type     = analysis,
    time     = 06,
    date     = 19970200,
    levtype  = model level,
    param    = 130.128,
    levelist = 1/to/137, 
    target   = "monthly"

Retrieve the temperature on model levels 1, 2 and 3 for a week old 12 hour forecast. > see in MARS Catalogue

retrieve,
    type    = fc,
    date    = -7,
    time    = 12,
    step    = 12,
    levtype = ml,
    level   = 1/2/3,
    param   = 130,
    target  = "t_ml_1_2_3"

Time and step

Retrieve the 1000 hPa geopotential from two weeks old 12 UTC forecast (> see in MARS Catalogue) for:

• the 12 hour forecast

  retrieve, type = fc, date = -14, time = 12, step = 12, levtype = pl, level = 1000, param = 129.128, target = "z1000_a"

  
  

• the 6, 12, 24 and 48 hour forecasts

  retrieve, type = fc, date = -14, time = 12, step = 6/12/24/48, levtype = pl, level = 1000, param = 129.128, target = "z1000_b"

  
  

• the 12 hour to 240 hour forecasts at 12 hour intervals

  retrieve, type = fc, date = -14, time = 12, step = 12/to/240/by/12, levtype = pl, level = 1000, param = 129.128, target = "z1000_c"

  
  

• all forecast time-steps (tip: you can use MARS catalogue to generate list of all available steps for you)

  retrieve, type = fc, date = -1, time = 12, # step = all, # The value "all" is convenient to use but MARS can only report on missing steps if # the full list of steps is specified: step = 0/1/2/3/4/5/6/7/8/9/10/11/12/13/14/15/16/17/18/19/20/21/22/23/24/25/26/27/28/29/30/31/32/33/34/35/36/37/38/39/40/41/42/43/44/45/46/47/48/49/50/51/52/53/54/55/56/57/58/59/60/61/62/63/64/65/66/67/68/69/70/71/72/73/74/75/76/77/78/79/80/81/82/83/84/85/86/87/88/89/90/93/96/99/102/105/108/111/114/117/120/123/126/129/132/135/138/141/144/150/156/162/168/174/180/186/192/198/204/210/216/222/228/234/240, levtype = pl, level = 1000, param = 129.128, target = "z1000_d"

  
  

ENS Overlap

Retrieve Total Precipitation, perturbed forecast overlap for 26th June 2023 (> see in MARS Catalogue)

retrieve,
    class = od,
    date = 2023-06-26,
    expver = 1,
    levtype = sfc,
    number = 1/to/50,
    param = 228.128,
    step = 336,
    stream = efov,
    time = 00:00:00,
    type = pf,
    target = [stream]_[date].grb

See also FAQ: "How can I compute the accumulation using data from the overlap stream?"

Post-processing

Retrieve the 1000,500 and 300 hPa temperatures valid for day before yesterday at noon from the 12 UTC forecast of four days ago on an N48 Gaussian Grid  > see in MARS Catalogue

retrieve, 
    class    = od,
    stream   = oper,
    expver   = 1,
    type     = forecast, 
    param    = 130.128, 
    date     = -4,
    time     = 12, 
    step     = 48, 
    grid     = N48, 
    levelist = 1000/500/300, 
    levtype  = pl, 
    target   = "t.gg.grb"

Retrieve the same on a latitude/longitude grid, 1.5 by 1.5 degrees. > see in MARS Catalogue

retrieve,
      class = od,
      stream = oper,
      expver = 1, 
      type = forecast,
      param = 130.128,
      date = -4,
      time = 12,
      step = 48,
      grid = 1.5/1.5,
      levelist = 1000/500/300,
      levtype = pl,
      target   = "t.ll.grb"

Retrieve Europe in the same way as in the previous retrieval. Compare the size of the resulting file with the previous one. > see in MARS Catalogue

retrieve,
    class = od,
    stream = oper,
    expver = 1, 
    type = forecast,
    param = 130.128,
    date = -4,
    time = 12,
    step = 48,
    grid = 1.5/1.5,
    levelist = 1000/500/300,
    levtype = pl, 
    area     = e,
    target   = "t.europe"

Retrieve the 1000 and 500 hPa geopotential for 1200 UTC analysis from two days ago in the formats below (one job for each form): > see in MARS Catalogue

• leaving the data in its original form

  retrieve, class = od, date = -2, expver = 1, levelist = 500/1000, levtype = pressure level, param = 129.128, stream = oper, time = 12, type=an,   target = "myfile_d"

  
  

• reducing the truncation to T106

  retrieve, class = od, date = -2, expver = 1, levelist = 500/1000, levtype = pressure level, param = 129.128, stream = oper, time = 12, type=an,    truncation = 106, target = "myfile_106"

  
  

• converting the data to a global N48 Gaussian grid

  retrieve, class = od, date = -2, expver = 1, levelist = 500/1000, levtype = pressure level, param = 129.128, stream = oper, time = 12, type=an, grid = N48, target = "myfile_N48"

  
  

• converting the data to a regular 5.0 by 5.0 degree global latitude/longitude grid

  retrieve, class = od, date = -2, expver = 1, levelist = 500/1000, levtype = pressure level, param = 129.128, stream = oper, time = 12, type=an, grid = 0.5/0.5, target = "myfile_ll"

  
  

• converting the data to an N80 Gaussian grid for the European area

  retrieve, class = od, date = -2, expver = 1, levelist = 500/1000, levtype = pressure level, param = 129.128, stream = oper, time = 12, type=an, grid = N80, target = "myfile_N80"

  
  

• converting the data to an N80 Gaussian grid for the area from 50 degrees North to 10 degrees North, and 20 degrees West to 20 degrees East

  retrieve, class = od, date = -2, expver = 1, levelist = 500/1000, levtype = pressure level, param = 129.128, stream = oper, time = 12, type = an, grid = N80, area = 50/-20/10/20, target = "myfile_N80_reg"

  
  

• converting the data to the same area but convert it to a 2.5 degree by 2.5 degree latitude/longitude grid

  retrieve, class = od, date = -2, expver = 1, levelist = 500/1000, levtype = pressure level, param = 129.128, stream = oper, time = 12, type = an, grid = 0.25/0.25, area = 50/-20/10/20, target = "myfile_ll_reg"

  
  

Retrieve surface winds for previous week's 24 hour forecast on the European area converting to a 0.57 by 0.57 degrees latitude/longitude grid. > see in MARS Catalogue
What happens to the conversion?

retrieve,
    class=od,
    date=-7,
    expver=1,
    levtype=sfc,
    param=165.128/166.128,
    step=24,
    stream=oper,
    time=12:00:00,
    type=fc,    
    area = europe,         # As the area is not compatible with the grid, 
    grid = 0.57/0.57,      # MARS will expand the area to make it fit.
    target = "10u_10v"

Observations

Retrieve all the sea surface observations for the 1st February 2015.  > see in MARS Catalogue

retrieve, 
    type    = ob, 
    obsgroup= con,
    obstype = ssd,
#   obstype = 9/11/12/13/14/19/21/22/23/180/181/182,

    date    = 20150201, 
    time    = 00,
    range   = 1439,
    target  = "obs.bufr"

Ensemble forecasts

Retrieve the 2 metre temperature field of the first 10 members of a week old 12 UTC ensemble prediction (type: perturbed forecast) for 2, 3 and 4 days in advance > see in MARS Catalogue

retrieve,
    class=od,
    date=-7,
    expver=1,
    levtype=sfc,
    number=1/to/10,
    param=167.128,
    step=48/to/96/by/24,
    stream=enfo,
    time=12,
    type=pf,
    target = "ensemble"

Retrieve the same for the control forecast. > see in MARS Catalogue

retrieve,
    class=od,
    date=-7,
    expver=1,
    levtype=sfc,
    param=167.128,
    step=48/to/96/by/24,
    stream=enfo,
    time=12,
    type=cf,
    target = "ensemble.cf"

ERA 5

Retrieve the vertical integral of eastward and northward heat flux as monthly means of daily means on a 1.125 by 1.125 degree latitude/longitude grid over Europe from the ERA 5 archive for January 2021 > see in MARS Catalogue

retrieve,
    class = ea,
    date = 20210101,
    decade = 2020,
    expver = 1,
    levtype = sfc,
    param = 69.162/70.162,
    stream = moda,
    type = an,
    area = europe,
    grid = 1.125/1.125, 
    target = "ea_moda_fluxes.grb"

Retrieval of snow depth from the ERA 5 archive for November 1993, for all analysis base times > see in MARS Catalogue. How many fields are requested?

retrieve,
  class   = ea,
  stream  = oper,
  expver  = 1,
  date    = 19931101/to/19931130,
  time    = 00/to/23,
  type    = an,
  levtype = sfc, 
  param   = 141.128,
  target  = "era5.199311.sd"

#                                  It retrieves 30 x 24 = 720 fields.

See also these Web API retrieve examples for ERA-Interim, S2S, TIGGE, ERA-20C etc.