A study evaluated differences between respirable dust concentration me
asured with three currently available cyclone samplers in a field sett
ing. Comparisons also were made between cyclone and personal cascade i
mpactor results. The cyclones selected were designed for sampling acco
rding to one of three respirable dust criteria: British Medical Resear
ch Council (BMRC) (D50 = 5 mum), previous American Conference of Gover
nmental Industrial Hygienists (ACGIH) (D50 = 3.5 mum), and proposed AC
GIH (D50 = 4 mum). Samples were collected by placing two each of three
brands of cyclone, two personal cascade impactors, and two total dust
samplers on a tripod-mounted boom assembly. The apparatus sampled abr
asive blasting operations. A significant difference was detected betwe
en the results of the BMRC and proposed ACGIH-type samplers and betwee
n the BMRC and previous ACGIH-type samplers. The BMRC-type sampler con
sistently yielded the highest average respirable dust concentrations,
followed by the proposed ACGIH and previous ACGIH-type samplers. Casca
de impactor results were converted to respirable dust concentrations u
sing the appropriate collection efficiencies for comparison to cyclone
results. The differences between the cyclone and cascade impactor res
ults were not statistically significant at the 5% level. Results reite
rate that important differences exist between respirable dust sampling
methods, and that when sample results are reported, the type of sampl
er used, the flow rate, and the respirable criteria with which the sam
pler was designed to conform should be specified.