Theoretical background: High levels of cash holdings allow companies to be financially flexible. Cash holders can react to business opportunities at once, and avoid dependence on external financing. Having cash reserves acts as a protective cushion from possible business shocks. However, cash holdings, like other asset items, incur cost. The direct costs mostly relate to obtaining and servicing of financing. Excessive cash may also induce the agency problem of free cash flow. Information asymmetry, among other factors, influences cost of capital and the efficiency of investor monitoring. Earnings management strategies are related to the informativeness of financial reports, as they can either support or disturb the assessment of company risk and performance. Hence, companies search for tailor-made strategies that usefully trade-off the benefits and costs of cash reserves and earnings quality. Purpose of the article: Earnings management occurs either via intervention in the reporting process using accruals, or changing the structure or the timing of real business transactions. This study aims to assess the differences in accrual and real earnings management strategies between firms with relatively high and low cash reserves. Based on a sample of non-financial companies listed on the Warsaw Stock Exchange in Poland, this research provides evidence on the relationship between the level of cash holdings and earnings quality. Research methods: Earnings management is proxied with annual cross sectional regressions estimated in non-financial GICS sectors. The Jones and the modified Jones model allow to estimate expected accruals. Using the Roychowdhury approach, normal cash flow from operations, normal production costs, and normal discretionary expenses are estimated. These normal levels are compared with their actual value to proxy for accrual- and real-earnings management. Next, each company was assigned to one of two groups, depending on the level of cash holdings relative to the annual industry medians. Parametric and non-parametric tests were used to test statistical significance. Main findings: The empirical results show statistically significant differences in accrual and real earnings management between companies with relatively high and low cash reserves. Cash holders seem to be more likely to manage earnings via shifting accruals between periods. However, there is no evidence that any of the groups boosted earnings more. Unsigned earnings management proxies suggest lower informativeness of financial reports for companies with cash buffers. Cash holdings were positively related to activities aimed at reducing fixed costs per unit and savings on discretionary expenses. Summing up, this research shows that companies with greater information asymmetry related to poor earnings quality preferred to hold larger cash reserves.