In order to test the phototoxic potency of the two tetracyclines most frequently prescribed in Sweden, a double-blind cross-over study using a double-dummy technique with doxycycline 0.1 g twice daily, lymecycline 0.6 g twice daily, and placebo, was performed in 15 healthy volunteers. Drugs were given for 3 consecutive days, and on the third day volunteers were tested with 25, 50, 75 and 100 J/cm2 of artificial long-wave ultraviolet radiation (UVA), and assessed 6 h later for erythematous photoreactions. Objective readings were made using skin reflectance spectrophotometry. All three substances were tested in each individual at weekly intervals. Within 50, 75 and 100 J/cm2 of UVA, lymecycline showed a slight increase in erythema compared with placebo, but this was not significant (50 and 100 J/cm2), or was of low significance (75 J/cm2). However, with the same doses, doxycycline showed a substantial increase in erythema compared with placebo, which was highly significant. We conclude that doxycycline has a higher phototoxic potency than lymecycline, and this is in agreement with earlier in vitro experimental data. We recommend that therapy with doxycycline is avoided during summer-time, and during holidays in a sunny climate.