Six genotypes of lentil (Lens culinaris) were grown in pots in a glasshouse under natural photoperiods of 10-12 h, extended to an inductive photoperiod of 16 h by supplementary illumination from incandescent lamps. The illuminance of the supplementary light was varied by placing replicate pots at different distances from the lamps; it decreased geometrically with distance (i.e. down an illuminance gradient) from about 2000 lux directly beneath the lamps to about 40 lux at a distance of 6 m from them. All genotypes came into flower progressively later as illuminance decreased; those selected in northern temperate regions were especially sensitive, taking about twice as long to flower at an illuminance of 40 compared with 2000 lux. We suggest, however, that these quantitative responses do not reflect differences in sensitivity to illuminance per se, but are the consequences of genotypic differences in sensitivity to photoperiod. The implications of these data are discussed in terms of adaptation to the aerial environment and strategies for screening germplasm and plant breeding in lentils and other grain legumes.Photoperiodically-sensitive plants flower seasonally in response to a particular duration of night. In natural environments the transition to and from darkness is not abrupt, but occurs through a gradually changing intensity of twilight. At what point does the plant begin to respond to darkness, or to light at dawn? Vince-Prue (1975) has considered these questions for plants in general but, to our knowledge (Summerfield, 1981), there are no data on these types of responses in lentils (Lens culinaris L.).Photoperiod is the only environmental factor which changes in an exactly predictable manner at any given latitude throughout every calendar year. These seasonal changes are a powerful determinant of phenological events in many crops of major economic importance, perhaps the best known of which are rice (Vergara and Chang, 1976) and soyabeans (Summerfield and Roberts, 1983a). Those lentil genotypes which respond to differences in photoperiod by coming into flower relatively early or late are described as 'quantitative long-day plants' (they will flower in a wide range of day-lengths but the onset of flowering is especially rapid in longer (16-24 h) than in shorter (6-12 h) days; Summerfield,