Objectives: The aim of this study was to examine non-cavitated approximal caries using non-invasive treatment methods. Materials and methods: Molar and premolar teeth with approximal caries were used in this in vivo study. Approximal caries lesions were evaluated with visual and radiographic inspection and with the DIAGNOdent device. Five groups were formed to study non-invasive treatment, and each had at least 25 early approximal carious lesions. Patients in the control group were not treated. After the separation, either ozone application, acidulated phosphate fluoride gel, CPP-ACP-containing material (Tooth Mousse), or an antibacterial bonding agent (Clearfil Protect Bond) was used. For 18 months after the non-invasive treatment, radiological controls were used to observe the progress of the initial and approximal caries in the 1st, 3rd, 6th and 12th months of follow-up. A Mann–Whitney U-test was used to perform the statistical analysis; in-group comparisons were made with the Wilcoxon signed-rank test, and a quantitative assessment was performed using a chi-squared test. Results: At the end of 18 months, the caries lesions in the control group were observed to progress (P < 0.01). The lesions that were scored as 1a during a visual inspection recovered by using non-invasive treatments. Conclusion: Approximal caries lesions that were detected at the early stages remained stationary when using antibacterial agents and materials that promoted remineralisation. Clinical relevance: Antibacterial agents and remineralisation materials can be used in treatment of early approximal caries lesions.