Environmental dependence of male sterility may sometimes result in residual fertility under ‘sterility conditions’, causing hybrid-seed contamination risk. An experiment was conducted to assess the risk factor and methods to increase hybrid-seed purity in a thermosensitive male-sterile tomato mutant, T-4, whose fertility is partially restored in autumn, but largely remains sterile in spring, with some residual fertility. Examination of pollen germination and the subsequent pollen-tube growth in vitro and on stigma revealed that a small proportion of the T-4 pollen was viable, with 10–20% germination, while normal pollen from ‘Tiny Tim’ had 60–85% germination 3–6 h after pollination. A stable male-sterile mutant T-3, whose pollen development collapses at the microspore stage, was pollinated with T-4 pollen followed by normal ‘Tiny Tim’ pollen with time lags of 2, 4, and 8 h. Concurrently, the T-4 mutant was self-pollinated by hand followed by normal pollen from an inbred line (M) with time lags of 2, 4, 8, 24, and 48 h. The progeny was scored for contamination based on differences in leaf characteristics. The percentage of T-4 seedlings (narrow leaved) in the F1 progeny was lowest at 2 h (0.2–6.3%),highest at 8 h (16.9–17.7%) and declined at 24 h–48 h (13.5–10.3%) time lag. The contamination rate was extremely low when pollination was done with normal pollen at (0.4%) and 24 h after (1.4%) anthesis without prior hand pollination with T-4 pollen. It was concluded that with pollination timing soon after anthesis, the T-4 mutant could be effectively applied in a two-line hybrid-seed production system with lower roguing cost of undesirable seedlings as opposed to the conventional three-line system.