The ability of the synthetic retinoid N-(4-hydroxyphenyl)retinamide (4-HPR) to affect the outcome of previously resected superficial bladder cancer was investigated in a pilot study using DNA content flow cytometry and conventional cytology as intermediate endpoints. Twelve patients were treated with oral 4-HPR (200 mg daily) and compared with 17 non-randomized, untreated controls. The median interval between transurethral resection and 4-HPR administration was 5.5 months (range 0-36). The median follow-up period was 12 months (range 3-31) in the 4-HPR group and 9 months (range 2-22) in the control group. The proportion of patients with DNA aneuploid stemlines in bladder-washed cells decreased from 7/12 (58%) to 5/11 (45%) in the 4-HPR group, but increased from 7/17 (41%) to 10/17 (59%) in the control group. In patients with stable diploid profiles, mean (+/- SE) S-phase and G2+M-phase fractions decreased in the course of retinoid treatment from basal levels of 15.2 +/- 4.1% to 7.5 +/- 3.3% and 10.3 +/- 2.2% to 5.2 +/- 0.4%, respectively. The same parameters in the control group changed from basal levels of 14.6 +/- 3.4% to 12.4 +/- 2.7% and 9.8 +/- 1.6% to 12.6 +/- 1.6%, respectively. Positive or suspicious cytologic examinations were present in 3/12 (25%) treated cases prior to 4-HPR administration and all subsequently reverted to normal. The same parameter in the control group increased from 4/17 (24%) to 6/17 (35%) during follow-up. Impaired adaptation to darkness was recorded in 4 patients, and transient dermatologic alterations were observed in one-third of the patients, requiring dose reduction in one case.(ABSTRACT TRUNCATED AT 250 WORDS)