The aim of the experiments was to find ways of increasing the yield of small molecular weight inhibitors of cell proliferation released by granulocytes. Almost pure populations of granulocytes from pig or human blood, or from sterile inflammatory exudates in rats were treated in various ways and then spun down. Molecules below approximately 10 000 dalton (Diaflo ultrafiltration or Sephadex G 25 filtration) in the supernatants were tested for inhibitory activity by measuring 3H-thymidine incorporation in 5 to 6-h coverslip cultures of rat bone marrow cells. The different granulocyte treatments were: Freeze-thawing, sonication, incubation (at +4 degrees -37 degrees C) in hypotonic media (0-200 mosm/kg), storage in vitro overnight (at +4 degrees C) before incubation, incubation at 37 degrees C in complete and buffered tissue culture medium (Fischer's with 10 mmol/1 HEPES), incubation in saline only (2-h periods, approximately 70 X 10(6) cells/ml), or with lidocaine added, with Ca++ and the Ca++ ionophore A-23187, with K+ and the K+ ionophore Valinomycin, with a high K+ concentration (50 mmol/1), with arachidonic acid, with a cAMP analogue, or with a protease inhibitor added during or at the end of the incubation. On a per cell basis rat peritonitis granulocytes released more inhibitor than pig blood granulocytes, whereas human blood granulocytes were not detectably inhibitory at all. Arachidonic acid was the most promising agent tested to increase inhibitor release above that occurring spontaneously from granulocytes incubated in saline.