The FHIT gene, located at the FRA3B fragile site of chromosome 3p14.2, encodes a 16.8 kD homologue of the yeast enzyme diadenosine tetraphosphate (Ap(4)A) hydrolase. Frequent allelic losses at this region in various malignancies, including non-small cell lung carcinomas (NSCLCs), imply that FHIT may represent a tumour suppressor gene (TSG). Increasing evidence suggests that multiple TSG impairment has a synergistic effect on tumour growth. The present study of 67 NSCLCs investigated the allelic imbalance (AIm) within the FHIT locus and its relationship with p53 abnormalities, kinetic parameters [proliferative activity or proliferation index (PI) and apoptotic index (AI)], and ploidy status of the carcinomas. Allelic imbalance at FHIT was observed in 35 out of 55 informative (heterozygous: H) cases (64%). Similar frequencies of loss of heterozygosity (LOH) were noticed among squamous cell lung carcinomas and adenocarcinomas. The high percentage of AIm in stage I tumours (71%) is indicative of its relatively early involvement in NSCL carcinogenesis. No association was found between LOH at FHIT, kinetic parameters, and ploidy status of the tumours. Concurrent loss at FHIT and p53 overexpression [FHIT(LOH)/p53(P)] was the most frequent pattern and was observed in 39% of the informative cases. The latter pattern was not associated with smoking, supporting the hypothesis that in patients with a history of tobacco exposure, FHIT allelic loss may not be a consequence of p53 checkpoint defects, but the outcome of tobacco-induced mutagenesis. Statistically significant differences in the presence of FHIT(LOH)/p53(P) and FHIT(LOH)/p53(N) patterns were noted at the proliferative and apoptotic level, whereas ploidy was similar amongst all groups, implying that wild-type (wt) p53 may play a safeguard role against altered FHIT function. However, the possibility of a masking effect from wt p53 cannot be excluded, since the FHIT(LOH)/p53(P) profile demonstrated a higher growth index (GI=PI/AI mean value ratio) than FHIT(H)/p53(P) (32 vs. 8), although this was not significant. Further studies are needed in order to elucidate the role of FHIT and its relationships with other cell-cycle regulatory molecules involved in NSCL carcinogenesis.