The paper studies price discrimination in price paths over first and subsequent purchase in settings where consumers have uncertainty over their valuation distributions for the product at the time of first purchase and realize their valuation after first purchase by incurring a learning cost that is different from price. Consumers pay (get paid or receive free) the product to try in the first period, incur an additional learning cost to realize their valuation before deciding to purchase the product in the second period. We propose and solve a sequential screening model based on dynamic mechanism design where forward looking consumers are offered an optimal menu of price paths over the two periods that price discriminates over the valuation uncertainty and learning costs. Price discrimination is feasible when learning costs are higher for the high valuation uncertainty segment. The firm offers a relatively lower first period price (including free or subsidies) for the higher learning cost-higher uncertainty segment, but a higher second period price; overall their total cost across both periods are higher. More consumers are excluded from second period purchase among the higher valuation uncertainty segment--as the firm seeks to obtain greater profits from those with the highest valuations from the higher uncertainty segment. Our model offers a dynamic screening explanation for the use of freemiums, even in settings without cross-customer externalities. We find that online reviews that reduce the variance in prior valuation distributions can dampen seller profits and the ability to offer upgrades in the second period can enhance the ability to price discriminate on valuation uncertainty more effectively and increase seller profits.