Objective: The aim of the study is to assess the language characteristics of plain language summaries (PLSs) of systematic reviews of oncology interventions in comparison to the language of corresponding scientific abstracts. Material and methods: In this cross sectional study we included all Cochrane scientific abstracts (SA) and corresponding PLSs of systematic reviews of oncology interventions available in the Cochrane up to February 2019, including Breast Cancer group; Childhood Cancer group; Gynaecological, Neuro-oncology and Orphan Cancers group; Haematology group; and Lung Cancer group. Language characteristics of PLS included text readability, measured using the SMOG index; and prevalence of words related to different language tones (clout, authenticity, emotions and analytical) measured using LIWC software. Conclusiveness of the summary was categorized into one of possible nine categories: positive, positive inconclusive, no evidence, no opinion, negative, negative inconclusive, unclear, equal and equal inconclusive. Results: In total, we collected 275 PLSs and corresponding SAs of systematic reviews of oncology interventions. In general, SMOG index of PLSs was slightly above the recommended 12 years of education for health information materials and the readability did not differ across Cochrane review groups. SMOG index of scientific abstracts was higher than readability index in corresponding PLSs. In general, the PLSs from the Colorectal Cancer group were the shortest, whereas the PLSs from the Haematological Malignancies group had the lowest proportion of words reflecting emotional tone. Analysis of the scientific abstracts followed the results from analysis of plain language summaries. In five out of six Cochrane groups, “Unclear“ category of conclusiveness was the most common one. Haematological Malignancies Group had the greatest proportion of „Positive“ conclusion category, but also the greatest number of '“Unclear“ conclusions. Childhood Cancer Group had the most “No evidence“ conclusion category. Analysis of the scientific abstracts found that conclusiveness was similar as in plain language summaries. Overall, PLSs with “No evidence” conclusion had the lowest SMOG index, as well as the fewest number of words compared to other conclusiveness categories. PLSs with “ Positive“ and “ Equal“ conclusiveness had the lowest proportion of words with analytical tone, while the PLSs with “No evidence” for any definite conclusion had a greater proportion of words related to clout tone. Conclusion: PLSs of Cochrane systematic reviews of oncological interventions have low readability and low emotional tone, as well as conclusiveness of the review findings. We intend to further compare PLSs and SAs, to understand the differences in the language of scientific and popular health information texts in order to suggest possible ways of improving the usability of information from Cochrane systematic reviews to the public.