The microvariability in AGNs was firstly announced and received with skepticism at the beginning of the 1960s, when a few sources were studied with a single-channel photoelectric photometer and the biggest Palomar telescopes. The availability and utilization of CCD cameras breathed new life into small telescopes. For many decades, the variability time-scales from minutes to years have been studied for many blazars using CCD cameras and small telescopes. In the last 30 years, we have been conducting a systematic study of over seventy blazars at the Abastumani Observatory on Mt. Kanobili using CCD cameras and two dedicated medium-sized 125 cm RCh and 70 cm meniscus telescopes. The observations were also conducted using the Calar-Alto 123 cm telescope.

From the middle of the 1990, over 4500 observing nights, we collected in BVRI bands about 600000 images. Among the studied blazars are thirty Northern Sky TeV sources such as Mrk421, Mrk 501, 1ES 0229+200, 1ES 0502+675, 1ES 0647+250, 1ES 0806+52.4, PG 1553+113, 1ES 1959+650, 1ES 2344+51.4, etc., and well-known classical sources such as BL Lacertae, 3C 454.4, 3C 279, S5 0716+714, CTA 102, AO 0235+164, 3C 66A, OJ 287, S4 0954+658, ON 231, 3C 273, PKS 1510-089, and 3C 345.

All images were reduced in MIDAS using DaoPhot II. The typical differential photometric accuracy for an R=15.0 star-like source in a 180 sec exposure is around one percent. A maximum variation for TeV sources up to 1.5 magnitude in the R band was detected, while a few other classical sources showed variations over five magnitudes. Pronounced microvariability at the level of 5% was detected in the case of BL Lacertae and S5 0716+714.