Background: The microleakage of a self-adhesive composite, a glass ionomer fissure sealant and a conventional resin-based fissure sealant were compared. Materials and methods: Fifty intact human molars with well-delineated pits and fissures were used and divided into 5 groups (n = 10). Group 1 specimens were etched (37% phosphoric acid) and sealed with conventional resin-based sealant (Helioseal F, Ivoclar Vivadent). Both Group 2 and 3 specimens were sealed with self-adhesive composite (Constic, DMG), but in Group 3, specimens were also etched (37% phosphoric acid). In Groups 4 and 5, specimens were sealed with a GIC sealant (Equia Fill, GC Company), but Group 5 was also exposed to thermo-light curing (TLC) with a LED polymerization unit (60 s). Subsequently, specimens were thermocycled (1800 cycles, dwelling time of 10 s), immersed in 50% silver nitrate solution (45 min), placed in a photo-developing solution (Kodak) under a lamp (120 W, 6 h), and cut into 3–4 slices. Marginal leakage (dye penetration depth) was evaluated under a light microscope and the worst score of each specimen was recorded (0–3). Results: Constic showed the lowest microleakage (Constic: 80% scored 0 or 1), followed by Helioseal (30% scored 0 or 1) (p = 0.037). Microleakage in groups sealed with Constic (with and without etching) were not different (p = 0.473). The quality of seal deteriorated after etching when Constic was used. However, TLC improved the seal when GIC sealant was used (p = 0.016) and also in comparison to Helioseal (p = 0.004). The TLC GIC sealant (Equia Fill, 90% scored 0 or 1) performed well, similar to self-adhesive composite (Constic, 80% scored 0 or 1) (p = 0.206). Conclusion: The present findings suggest that the self-adhesive sealant and the GIC sealant that were exposed to TLC had comparable sealing ability and superior sealing characteristics compared to the conventional resin-based sealant. A long-term clinical trial is needed to assess the intra-oral performance.