Seeds of the narrow-endemic Solidago shortii and of the geographically-widespread S. altissima and S. nemoralis collected and buried in 1992 and 1993 were incubated in light and in darkness at 15/6°, 20/10°, 25/15°, 30/15° or 35/20°C following various periods of burial in soil in a non-temperature-controlled glasshouse. At maturity in November, seeds of the three species germinated to 0–1% in light at 15/6°C and to 10–77% at 20/10°, 25/15°, 30/15° and 35/20°C. Seeds exhumed each April from 1993 to 1996 and incubated in light at 15/6° and 20/10°C germinated to ≥83% and ≥90%, respectively, whereas those exhumed each September of 1993–96 germinated to ≤2%and ≤40%, respectively. At 25/15°, 30/15° and 35/20°C in light, seeds of S. altissima and S. shortii germinated to ≥52% and those of S. nemoralis to ≥19/, regardless of when they were exhumed. Timson's index, which integrates percentages, rates and times for onset of germination, was higher at all temperature regimes for seeds exhumed in April 1995 than for those exhumed in September 1995. Freshly-matured seeds of the three species germinated to 0–11% in darkness. Furthermore, regardless of when they were exhumed, seeds of S. altissima and S. nemoralis incubated in darkness germinated mostly to only 0–9% over the range of temperature regimes. In contrast, ≤88% and ≤6% of the seeds of S. shortii exhumed and incubated in darkness each April and September of 1993–96, respectively, germinated, ≤1% of them germinating while buried in soil. Thus, although buried seeds of all three species exhibited an annual conditional dormancy/non-dormancy cycle, only those of S. shortii exhibited cyclic changes in their germination response in darkness.