Though the notion of the aspectual pair is certainly useful, there is a strong sense that pairedness is not a uniform phenomenon, and neither are perfectivization and imperfectivization. This paper will propose a new model of the derivational relations between corresponding perfective and imperfective verbs that accounts for varying types of perfectives and imperfectives and links between them. This model consists of a conceptual Action Space and traces the paths through this space constructed via perfective and imperfective derivation. This model draws on several theoretical sources, particularly: Croft’s (Croft, William. 2000. Radical Construction Grammar. Oxford: Oxford UP) claim that conceptual space is universal and that language-specific conventions map varying divisions onto conceptual space; Smith’s (Smith, Carlota. 1991. The Parameter of Aspect. Dordrecht: Kluwer) claim that situation aspect (= universal concepts) is covert, whereas viewpoint aspect (= language-specific concepts) is overt (= morphologically marked); Mehlig’s (articles dated 1994, 1996, 2003) classification of Russian verbs as heterogeneous vs. homogeneous, transformative vs. non, singulative, rehomogenized, etc.; and Janda’s (forthcoming article) metaphorical model where PERFECTIVE IS A DISCRETE SOLID and IMPERFECTIVE IS A FLUID SUBSTANCE. On the basis of these ideas, I posit a conceptual space of aspect, with a central zone of processes (rabotat′), repetitions (podpisyvat′ zakony, prygat′) and completable activities (stroit′ dom), surrounded by Acts which can be Completion Acts the culmination of a completable activity, with three further distinctions), Complex Acts (a combination of a solid boundary with a fluid activity: porabotat′, popodpisyvat′, poprygat′) and Single Acts (prygnut′). Completion Acts are further distinguished as Predictable (postroit′ dom) vs. Unpredictable (ugovorit′ podrugu – corresponding imperfective interpretable as conative), Irreversible (sšit′ plat′e) vs. Reversible (otkryt′ okno, leč′ – corresponding imperfective interpretable as annulled action), and Instantaneous (uznat′ znakomogo) vs. Non-Instantaneous (pročitat′ knigu). In Russian there are firm boundaries between Activities and Acts and between the various Acts, requiring the derivation of a new verb in order to cross these boundaries. These derivations navigate the conceptual space along many different paths. Each path connects two endpoints, motivating the notion of pairedness, but many paths are possible, and verbs can participate in aggregated paths. For example: Activity > Completion Act > Activity > Complex Act, as realized by krasit′ > pokrasit′ > pokrašivat′ > popokrašivat′. Or: Activity – Single Act >> Completion Act > Activity > Complex Act, as realized by prygat′ – prygnut′ >> pereprygnut′ > pereprygivat′ > popereprygivat′. This model provides a much richer account of the relations of verbs created by aspectual derivation than mere pairedness, and I will explore important implications for motion verbs, historical change, and the semantics of aspect.